Wednesday, 07 November 2001


Published in Analytical Articles

By S. Frederick Starr (11/7/2001 issue of the CACI Analyst)

BAKGROUND: Like all its neighbors, including Iran and Pakistan, Afghanistan is a multi-ethnic land.  Pashtuns are the largest group and Tajiks a distant second, followed by Hazaras, Uzbeks, and Turkmens.  Obviously, any government that is to be legitimate must be organized in such a way as to assure fair representation for all.

BAKGROUND: Like all its neighbors, including Iran and Pakistan, Afghanistan is a multi-ethnic land.  Pashtuns are the largest group and Tajiks a distant second, followed by Hazaras, Uzbeks, and Turkmens.  Obviously, any government that is to be legitimate must be organized in such a way as to assure fair representation for all.  Every Afghan appreciates this, and knows that on this point the Taliban’s Pashtun leaders have failed miserably.  Foreign diplomats scurrying to plan the new government have also taken this need to heart.  But many have gone further. During a recent trip to Tajikistan, Germany’s Foreign Minister Joschka Fischer announced that the new government should be a federation.  Washington is also abuzz with talk of federalism for Afghanistan.  But the notion of a federalized Afghanistan is not only wrong-headed but dangerous. To be sure, advocates of a federalized Afghanistan can point to several obvious benefits of such an arrangement.  Everyone wants a say in his or her own government and, in a nation of separate ethnic groups, a federal system can assure such a voice.  It can also provide each territorial group with a decisive say over matters that concern only that one locality.  By doing so, it can foster participation at the regional level.  And in so doing, it can also create a counterbalance to what many fear could become an overweening central government dominated by just one group, which in Azfghanistan would be the Pashtuns. Federal systems are able to fulfill all these functions in Germany, Canada, Switzerland, or the United States. But such systems are peculiarly difficult to bring into being. The United States endured a bloody civil war before it achieved a sustainable balance between federal and state authority; Germany did not adopt a federal system until it lost World War II; Switzerland’s was hundreds of years in the making; and Quebec is still testing the limits of Canadian federalism. In each of these cases, an overriding concern of the founding fathers was to place firm limits on central authority.  They feared centripetal more than centrifugal forces. But in Afghanistan the gravest dangers arise from an excess of centrifugal forces. A glance at some of the country’s most urgent requirements reveals the overwhelming need for a functioning center.

IMPLICATIONS: Thus, Afghanistan’s new government must be able to assure that laws are enacted and fairly implemented throughout the land. It must provide for security. It must be capable of preventing unlawful movement across external borders, including the interdiction of terrorists and drug traffickers. It must assure that civil order is maintained on equitable terms across the land, so that criminals cannot take refuge by fleeing across internal borders. And it must be able to assure the smooth flow of trade—in practice, truck traffic—across the entire country, without which the economy will never move off dead center. Finally, it must be capable of distributing foreign aid in an equitable manner that leaves no room for claims of ethnic favoritism.  Each of these tasks will be rendered more difficult by a federal system. Not only will it enshrine centrifugal forces in law but it will give them endless opportunities to challenge the underpinnings of the state. Will there be a right of secession? Will ‘states’ have the power to levy their own taxes and duties on trade, and to set the rates? Will federal law prevail over all state laws? Can the ‘states’ make separate arrangements with foreign countries and donors in areas like trade and investment? How will inter-state conflicts be adjudicated? But such issues of principle may never arise, since a federal system is likely to die on the drawing board, literally. Afghanistan’s states will not exist until someone draws up a new set of borders. The existing provinces only vaguely reflect ethnic divisions. What rights will minority populations have in ethnic states and how will they be protected?  As Afghanistan’s Jeffersons and Madisons set to work on this problem, another one will arise as they encounter the ‘big state-small state’ problem that vexed America’s Framers in Philadelphia. Will the numerical preponderant Pashtuns be divided into several smaller states in order to create a semblance of equity among the federal units? Even if Pashtuns agree to this, which is unlikely, how will the new borders be drawn, and by whom? As a reborn and federalized Afghanistan grapples with these challenges, normal governmental functions will be in abeyance.  If this situation does not lead immediately to a renewal of warlordism, then the first efforts to make a federal system work will likely do so, by inviting the reemergence of leaders dedicated to promoting their own ethnic group against all others and at the expense of the very concept of national citizenship. The idea of giving each group its own ethnic sandbox may be attractive to a Green politician in Berlin or to a champion of states-rights from Arkansas. In Afghanistan it is a dangerous and irresponsible proposal.  Is there an alternative that meets both sets of needs defined above? Afghanistan needs a unitary government based on the existing twenty-nine provinces, with responsibility for security, police, customs, and all foreign relations vested firmly in the capital. It needs a strong and responsible central government and local administrators who are competent and honest. But it does not need a bureaucratized system, let alone one that, like that of the former emirs and kings, was dedicated mainly to extracting taxes from the people at any cost. Firm limits must be placed on the central authority, and especially on the power to impose taxes. Afghanistan needs a unitary system, but one that accords a substantial role to local initiative.  At the least, it should allow for the local election of governors—if not immediately, then after a brief transitional period.  And above all, the lowest level of the system should be grounded not in an army of local bureaucrats but in the existing institutions of self-government at the village, tribal, and district levels. Through two decades of war and chaos, age-old institutions of local self-government have continued to function in many areas.  Whether urban mahallas or guzars, kinship groups, or tribal entities, these have always been, and remain today, the primary focus of loyalty and even identity. As the Afghan-born scholar Nazif Shahrani has pointed out, these are the essential and unavoidable units in Afghan society and certainly the only ones that enjoy widespread credibility. 

CONCLUSIONS: At some future point it might be possible for Afghanistan, like Germany, to move to a federal system, and even a decentralized one. To attempt such an arrangement today would be to turn the country into a highly flammable laboratory of political science.  Instead, a unitary system should be adopted, but one that allows participation at the provincial and national level and which, at the most local level, is solidly rooted in communal self-government. In the end, the only government that will succeed in Afghanistan is one planned by Afghans themselves, and not by enthusiasts from abroad.  It is well and good for foreign friends to offer advice, as is being done here.  But Afghans should know that international experience does not all point in the direction of federalism, and, at the same time, that a bureaucratized and overly centralized system, no matter how efficient, cannot make up for the absence of citizen participation and for the active involvement of traditional entities at the most local level.

AUTHOR BIO: Prof. S. Frederick Starr is the Chairman of the Central Asia-Caucasus Institute of Johns Hopkins University-SAIS, and the rector pro tempore of the University of Central Asia.

Copyright 2001 The Analyst All rights reserved

Read 4936 times

Visit also





Staff Publications

Op-ed Mamuka Tsereteli and James Jay Carafano, Tsereteli & Carafano: Putin threatens Ukraine – here's the danger and what US, allies should do about it, Fox News, April 13, 2021.

Op-ed S. Enders Wimbish, US withdrawal from Afghanistan spells dangerous geopolitical realignments, The Hill, April 2, 2021.  

Silk Road Paper Svante E. Cornell and S. Frederick Starr, Kazakhstan's Role in International Mediation under First President Nursultan Nazarbayev, November 2020.

Book S. Frederick Starr, Eldar Ismailov, Nazim Muzaffarli, Basic Principles for the Rehabilitation of Azerbaijan’s Post-Conflict Territories, 2010.

Analysis Svante E. Cornell, How Did Armenia So Badly Miscalculate Its War with Azerbaijan? The National Interest, November 14, 2020.

Op-ed Svante E. Cornell, Halting the conflict between Armenia and Azerbaijan: Russian Peacekeeping is not the Solution Washington Times, October 20, 2020.

Analysis Svante E. Cornell, Can America Stop a Wider War between Armenia and Azerbaijan? The National Interest, October 5, 2020.

Article S. Frederick Starr, America Inches Toward a Serious Central Asia Strategy AFPC Defense Dossier, June 3, 2020.

Silk Road Paper Farrukh Irnazarov and Roman Vakulchuk, Discovering Opportunities in the Pandemic? Four Economic Response Scenarios for Central Asia, July 2020.  


The Central Asia-Caucasus Analyst is a biweekly publication of the Central Asia-Caucasus Institute & Silk Road Studies Program, a Joint Transatlantic Research and Policy Center affiliated with the American Foreign Policy Council, Washington DC., and the Institute for Security and Development Policy, Stockholm. For 15 years, the Analyst has brought cutting edge analysis of the region geared toward a practitioner audience.


Sign up for upcoming events, latest news and articles from the CACI Analyst